With all the buzz around ChatGPT, artificial intelligence (AI) is a hot topic right now across nearly every business sector and application. Project portfolio management is no different. Even Gartner has predicted that by 2030, AI will take over and “eliminate” as much as 80% of project management tasks, including data collection, tracking and reporting.
While the promise of generative AI (the likes of ChatGPT) may be a bit further off in this field, it’s true there are many possibilities for AI within PPM. But it’s also not without some significant challenges that must be overcome. Let’s talk about what it can do, but also some pitfalls and concerns you need to be aware of.
- Natural Language Processing (NLP) can be a game-changer for efficiency. If you’ve ever asked Alexa anything or said, “Hey Siri…”you’ve used NLP—a form of AI that can recognize and respond appropriately to natural spoken language. In the context of PPM, being able to query your system of record about specific needs can drastically improve data insights and save time. For just one example, instead of poring over Gantt charts and spreadsheets and comparing those to manpower management, you can simply ask the platform, “Do I have resources for [insert specific task or project] next week?” Or, “Who can fill in for James when he’s on holiday next week?” With NLP capabilities, the system can help you find resources with similar skills and availability to plan appropriately.
- Data collection and analysis. Gartner is right: the potential here is huge. After all, chasing down data to report to management is a huge part of the PPM function, and AI will give us the ability to do that much quicker, and arrive at performance insights and risk assessments faster than ever. Not only will this help guide decision-making, but it will enable PMs to do what they should be doing: coaching and supporting teams, instead of sifting through emails and Slack messages and scrolling through row after row of a spreadsheet or Gantt chart trying to find critical nuggets of data.
- Dynamic scenario planning. AI’s super-human ability to analyze data and deliver insights at incredible speed provides unprecedented clarity that can guide realistic scenario planning. Let’s say you’re planning next year’s project roadmap. You already have 50 projects in process, and you want to add 40 more. With AI-powered scenario planning, you can see, in real-time, whether you realistically have the resources to support that plan, whether you should put some projects on hold, which you should prioritize, and perhaps discover you’ll need to hire additional talent (or not, depending on your priorities, budget, etc.). This ability to flexibly develop and test different scenarios is transformative.
- Accelerate decision making. Because AI enables these real-time insights, an AI-driven PPM platform allows you to see various parameters, dependencies, and roadblocks more clearly so that you can make better-informed decisions and adapt on the fly. Prior to AI, in that same meeting from above where you’re planning next year’s roadmap, you’d have to discuss potential scenarios, adjourn the meeting, go back to the drawing board to work on revised plans, and then reconvene in three or four days to review the options. AI can dramatically accelerate that process, so you can stay on track with timelines and business objectives, not to mention gain or maintain a competitive advantage by being able to move faster.
- More insightful risk analysis. The ability to analyze more data from more sources, and do it faster, can provide better insight into risk potential. Especially as projects become more complex and involve cross-team coordination, there are a lot of factors that contribute to risk: the more tasks, people, and dependencies, the higher the risk of delays, resource shortfalls, missed OKRs, and ultimately, project failure. Using AI to analyze real time and historical data can help PPMs identify those roadblocks and risks sooner, and help you devise and navigate mitigation strategies.
While all of this sounds amazing—and some of this capability already exists in the market—there are a few areas of caution that organizations should consider before diving headfirst into the AI-driven solution pool.
- Don’t adopt AI for the sake of AI. Don’t look at AI as a tool you need—instead, decide what functions or capabilities you need first, and then find an AI tool that delivers. With so much hype around AI, it’s tempting to think you have to have it, but if it doesn’t serve your needs, it’s a waste of time and resources.
- Take an incremental approach. Going all-in on AI right out of the gate can be a mistake, completely upheaving your entire workflow and exposing potential shortcomings in the solution only after it’s been fully implemented. As with any new platform, a managed approach to adoption and implementation is essential. You’ll want to ease into the process to minimize risk, assure user confidence and understand what’s working and what’s not throughout the process.
- Prepare to standardize. One of the biggest problems in many organizations is scattered and siloed data, and that prevents them from gaining insights that depict an accurate and complete picture. An AI solution will have the same problem if you don’t standardize it as the system of record across the organization. If marketing is using one tool and product development is using another, an AI solution can’t magically integrate that data. It’s essential that everyone is on the same system in order to be on the same page.
- Beware of “black box” solutions. One of the biggest risks that also diminishes trust in AI solutions is that they don’t explain how and why they arrived at that solution. ChatGPT doesn’t cite sources. In a business context, users need to know where the data is coming from and how the AI solution arrived at its conclusions because there may be nuances that the platform doesn’t understand. Just like in middle school algebra, an AI platform should show its work, not just the final answer.
There’s no doubt AI has tremendous potential to improve the efficiency, operational workflow, clarity and accuracy of the project management function. And there are tools that are already implementing these capabilities and showing promise. For organizations looking to jump on the AI opportunity, a careful, results-oriented approach with phased adoption is the most prudent strategy.
